DOCSLIB.ORG

Lustrearchitecture-2017-08-17, V4

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Lustrearchitecture-2017-08-17, V4 Introduction to Lustre* Architecture Lustre* systems and network administration October 2017 * Other names and brands may be claimed as the property of others Lustre – Fast, Scalable Storage for HPC Lustre* is an open-source, object-based, distributed, parallel, clustered file system § Designed for maximum performance at massive scale § Capable of Exascale capacities § Highest IO performance available for the world’s largest supercomputers § POSIX compliant § Efficient and cost effective 3 Lustre* is an open-source, distributed, parallel data storage platform designed for massive scalability, high-performance, and high-availability. Popular with the HPC community, Lustre is used to support the most demanding data-intensive applications. Lustre provides horizontally scalable IO for data centres of all sizes, and is deployed alongside some of the very largest supercomputers. The majority of the top 100 fastest computers, as measured by top500.org, use Lustre for their high performance, scalable storage. Lustre file systems can scale from very small platforms of a few hundred terabytes up to large scale platforms with hundreds of petabytes, in a single, POSIX-compliant, name space. Capacity and throughput performance scale easily. Lustre runs on Linux-based operating systems and employs a client-server network architecture. Lustre software services are implemented entirely within the Linux kernel, as loadable modules. Storage is provided by a set of servers that can scale to populations measuring up to several hundred hosts. Lustre servers for a single file system instance can, in aggregate, present up to tens of petabytes of storage to thousands of compute clients simultaneously, and deliver more than a terabyte-per-second of combined throughput. Lustre Scalable Storage Management Metadata DNE Metadata Object Storage Target (MGT) Target (MDT0) Targets (MDTi - MDTj) Targets (OSTs) Storage servers grouped into failover pairs Management Management Additional Object Storage Object Storage Network & Metadata Metadata Servers Servers Servers Servers High Performance Data Network (Omni-Path, InfiniBand, Ethernet) Lustre Clients (1 – 100,000+) 4 Lustre’s architecture uses distributed, object-based storage managed by servers and accessed by client computers using an efficient network protocol. There are metadata servers, responsible for storage allocation, and managing the file system name space, and object storage servers, responsible for the data content itself. A file in Lustre is comprised of a metadata inode object and one or more data objects. Lustre is a client-server, parallel, distributed, network file system. Servers manage the presentation of storage to a network of clients, and write data sent from clients to persistent storage targets. There are three different classes of server: • Management server provides configuration information, file system registries • Metadata servers record file system namespace, inodes. The metadata servers maintain the file system index. • Object storage servers record file content in distributed binary objects. A single file is comprised of 1 or more objects, and the data for that file is organized in stripes across the objects. Objects are distributed across the available storage targets. Lustre separates metadata (inode) storage from block data storage (file content). All file metadata operations (creating and deleting files, allocating data objects, managing permissions) are managed by the metadata servers. Metadata servers provide the index to the file system. Metadata is stored in key-value index objects that store file system inodes: file and directory names, permissions, block data locations, extended attributes, etc. Lustre object storage servers write the data content out to persistent storage. Object servers can be written to concurrently, and individual files can be distributed across multiple objects on multiple servers. This allows very large files to be created and accessed in parallel by processes distributed across a network of computer infrastructure. Block data is stored in binary byte array data objects on a set of storage targets. A single Lustre "file" can be written to multiple objects across many storage targets. In addition to the metadata and object data, there is the management service, which is used to keep track of servers, clients, storage targets and file system configuration parameters. Clients aggregate the metadata name space and object data to present a coherent POSIX file system to applications. Clients do not access storage directly: all I/O is sent over a network. The Lustre client software splits IO operations into metadata and block data, communicating with the appropriate services to service IO transactions. This is the key concept of Lustre’s design – separate small, random, IOPS-intensive metadata traffic from the large, throughput- intensive, streaming block IO. Lustre Building Blocks Building block for Scalability. Management Metadata Object Storage Add more OSS Target (MGT) Target (MDT) Targets (OSTs) Metadata is stored building blocks to separately from increase capacity file object data. and throughput With DNE, multiple bandwidth. metadata servers can be added to Uniformity in building increase block configuration Storage servers grouped namespace promotes consistency into failover pairs capacity and in performance and performance behavior as the file system grows. Management Metadata Object Storage Network Servers Servers Balance server IO with storage IO Intel Manager for Lustre capability for best utilisation. High Performance Data Network (InfiniBand, 10GbE) Lustre Clients (1 – 100,000+) 5 The major components of a Lustre file system cluster are: • MGS + MGT: Management service, provides a registry of all active Lustre servers and clients, and stores Lustre configuration information. MGT is the management service storage target used to store configuration data. • MDS + MDTs: Metadata service, provides file system namespace (the file system index), storing the inodes for a file system. MDT is the metadata storage target, the storage device used to hold metadata information persistently. Multiple MDS and MDTs can be added to provide metadata scaling. • OSS + OSTs: Object Storage service, provides Bulk storage of data. Files can Be written in stripes across multiple object storage targets (OSTs). Striping delivers scalable performance and capacity for files. OSS are the primary scalable service unit that determines overall aggregate throughput and capacity of the file system. • Clients: Lustre clients mount each Lustre file system instance using the Lustre Network protocol (LNet). Presents a POSIX-compliant file system to the OS. Applications use standard POSIX system calls for Lustre IO, and do not need to be written specifically for Lustre. • Network: Lustre is a network-based file system, all IO transactions are sent using network RPCs. Clients have no local persistent storage and are often diskless. Supports many different network technologies, including OPA, IB, Ethernet. Lustre Networking Applications do not run directly on storage servers: all application I/O is transacted over a network. Lustre network I/O is transmitted using a protocol called LNet, derived from the Portals network programming interface. LNet has native support for TCP/IP networks as well as RDMA networks such as Intel Omni-Path Architecture (OPA) and InfiniBand. LNet supports heterogeneous network environments. LNet can aggregate IO across independent interfaces, enabling network multipathing. Servers and clients can be multi-homed, and traffic can be routed using dedicated machines called LNet routers. Lustre network (LNet) routers provide a gateway between different LNet networks. Multiple routers can be grouped into pools to provide performance scalability and to provide multiple routes for availability. The Lustre network protocol is connection-based: end-points maintain shared, coordinated state. Servers maintain exports for each active connection from a client to a server storage target, and clients maintain imports as an inverse of server exports. Connections are per- target and per-client: if a server exports N targets to P clients, there will be (N * P) exports on the server, and each client will have N imports from that server. Clients will have imports for every Lustre storage target from every server that represents the file system. Most Lustre protocol actions are initiated by clients. The most common activity in the Lustre protocol is for a client to initiate an RPC to a specific target. A server may also initiate an RPC to the target on another server, e.g. an MDS RPC to the MGS for configuration data; or an RPC from MDS to an OST to update the MDS’s state with available space data. Object storage servers never communicate with other object storage servers: all coordination is managed via the MDS or MGS. OSS do not initiate connections to clients or to an MDS. An OSS is relatively passive: it waits for incoming requests from either an MDS or Lustre clients. Lustre and Linux The core of Lustre runs in the Linux kernel on both servers and clients. Lustre servers have a choice of backend storage target formats, either LDISKFS (derived from EXT4), or ZFS (ported from OpenSolaris to Linux). Lustre servers using LDISKFS storage require patches to the Linux kernel. These patches are to improve performance, or to enable instrumentation useful during the automated test processes in Lustre’s software development lifecycle. The list of patches continues to reduce
Load more

Recommended publications
  • Oracle® ZFS Storage Appliance Security Guide, Release OS8.6.X
    ® Oracle ZFS Storage Appliance Security Guide, Release OS8.6.x Part No: E76480-01 September 2016 Oracle ZFS Storage Appliance Security Guide, Release OS8.6.x Part No: E76480-01 Copyright © 2014, 2016, Oracle and/or its affiliates. All rights reserved. This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is prohibited. The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please report them to us in writing. If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, then the following notice is applicable: U.S. GOVERNMENT END USERS. Oracle programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation, delivered to U.S. Government end users are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental regulations. As such, use, duplication, disclosure, modification, and adaptation of the programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation, shall be subject to license terms and license restrictions applicable to the programs.
    [Show full text]
  • Altavault 4.4 Administration Guide
    Beta Draft NetApp® AltaVault™ Cloud Integrated Storage 4.4 Administration Guide NetApp, Inc. Telephone: +1 (408) 822-6000 Part number: 215-12478_A0 495 East Java Drive Fax: + 1 (408) 822-4501 November 2017 Sunnyvale, CA 94089 Support telephone: +1(888) 463-8277 U.S. Web: www.netapp.com Feedback: [email protected] Beta Draft Contents Beta Draft Contents Chapter 1 - Introduction of NetApp AltaVault Cloud Integrated Storage ............................................ 11 Overview of AltaVault....................................................................................................................................11 Supported backup applications and cloud destinations...........................................................................11 AutoSupport ............................................................................................................................................11 System requirements and specifications.........................................................................................................11 Documentation and release notes ...................................................................................................................12 Chapter 2 - Deploying the AltaVault appliance ......................................................................................13 Deployment guidelines ...................................................................................................................................13 Basic configuration.........................................................................................................................................15
    [Show full text]
  • The Parallel File System Lustre
    The parallel file system Lustre Roland Laifer STEINBUCH CENTRE FOR COMPUTING - SCC KIT – University of the State Rolandof Baden Laifer-Württemberg – Internal and SCC Storage Workshop National Laboratory of the Helmholtz Association www.kit.edu Overview Basic Lustre concepts Lustre status Vendors New features Pros and cons INSTITUTSLustre-, FAKULTÄTS systems-, ABTEILUNGSNAME at (inKIT der Masteransicht ändern) Complexity of underlying hardware Remarks on Lustre performance 2 16.4.2014 Roland Laifer – Internal SCC Storage Workshop Steinbuch Centre for Computing Basic Lustre concepts Client ClientClient Directory operations, file open/close File I/O & file locking metadata & concurrency INSTITUTS-, FAKULTÄTS-, ABTEILUNGSNAME (in der Recovery,Masteransicht ändern)file status, Metadata Server file creation Object Storage Server Lustre componets: Clients offer standard file system API (POSIX) Metadata servers (MDS) hold metadata, e.g. directory data, and store them on Metadata Targets (MDTs) Object Storage Servers (OSS) hold file contents and store them on Object Storage Targets (OSTs) All communicate efficiently over interconnects, e.g. with RDMA 3 16.4.2014 Roland Laifer – Internal SCC Storage Workshop Steinbuch Centre for Computing Lustre status (1) Huge user base about 70% of Top100 use Lustre Lustre HW + SW solutions available from many vendors: DDN (via resellers, e.g. HP, Dell), Xyratex – now Seagate (via resellers, e.g. Cray, HP), Bull, NEC, NetApp, EMC, SGI Lustre is Open Source INSTITUTS-, LotsFAKULTÄTS of organizational-, ABTEILUNGSNAME
    [Show full text]
  • A Fog Storage Software Architecture for the Internet of Things Bastien Confais, Adrien Lebre, Benoît Parrein
    A Fog storage software architecture for the Internet of Things Bastien Confais, Adrien Lebre, Benoît Parrein To cite this version: Bastien Confais, Adrien Lebre, Benoît Parrein. A Fog storage software architecture for the Internet of Things. Advances in Edge Computing: Massive Parallel Processing and Applications, IOS Press, pp.61-105, 2020, Advances in Parallel Computing, 978-1-64368-062-0. 10.3233/APC200004. hal- 02496105 HAL Id: hal-02496105 https://hal.archives-ouvertes.fr/hal-02496105 Submitted on 2 Mar 2020 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. November 2019 A Fog storage software architecture for the Internet of Things Bastien CONFAIS a Adrien LEBRE b and Benoˆıt PARREIN c;1 a CNRS, LS2N, Polytech Nantes, rue Christian Pauc, Nantes, France b Institut Mines Telecom Atlantique, LS2N/Inria, 4 Rue Alfred Kastler, Nantes, France c Universite´ de Nantes, LS2N, Polytech Nantes, Nantes, France Abstract. The last prevision of the european Think Tank IDATE Digiworld esti- mates to 35 billion of connected devices in 2030 over the world just for the con- sumer market. This deep wave will be accompanied by a deluge of data, applica- tions and services.
    [Show full text]
  • What Is Object Storage?
    What is Object Storage? What is object storage? How does object storage vs file system compare? When should object storage be used? This short paper looks at the technical side of why object storage is often a better building block for storage platforms than file systems are. www.object-matrix.com Object Matrix Ltd [email protected] Experts in Digital Content Governance & Object Storage +44(0)2920 382 308 The Rise of Object Storage Centera the trail blazer… What exactly Object Storage is made of will be discussed later; its benefits and its limitations included. But first of all a brief history of the rise of Object Storage: Concepts around object storage can be dated back to the 1980’s1 , but it wasn’t until around 2002 when EMC launched Centera to the world – a Content Addressable Storage product2 - that there was an object storage product for the world in general3. However, whilst Centera sold well – some sources say over 600PB were sold – there were fundamental issues with the product. In, 2005 I had a meeting with a “next Companies railed against having to use a “proprietary API” for data generation guru” of a top 3 storage access and a simple search on a search engine shows that company, and he boldly told me: “There is Centera had plenty of complaints about its performance. It wasn’t no place for Object Storage. Everything long until the industry was calling time on Centera and its “content you can do on object storage can be addressable storage” (CAS) version of object storage: not only done in the filesystem.
    [Show full text]
  • Softnas Deployment Guide for High- Performance SQL Storage
    SoftNAS Deployment Guide for High- Performance SQL Storage Introduction SoftNAS cloud NAS systems are based on an innovative, memory-centric storage architecture that delivers unparalleled NAS performance, efficiency, and value. They incorporate a hybrid disk storage technology that tailors the usage of data disks, log solid- state cache drives (SSDs), and read cache SSDs to the data share's specific needs. Additional features include variable storage record size, data compression, and multiple connectivity options. As a Cloud NAS solution, SoftNAS cloud NAS systems provide an excellent base for Microsoft Windows Server deployments by providing iSCSI or Fibre Channel block storage for Microsoft SQL Server, and network file system (NFS) or server message block (SMB) file storage for Microsoft Windows client access. This document covers the best practices to follow when deploying Microsoft SQL Server on a SoftNAS cloud NAS system. The intended audience is storage administrators and Microsoft SQL Server database administrators. Maintaining High Availability As with any business-critical application, high availability is a crucial design criterion to be considered when deploying a Microsoft SQL Server installation. Microsoft SQL Server 2016 can be installed on local and/or shared file systems, and SoftNAS cloud NAS systems can satisfy both of these options. Local file systems (from the Microsoft Windows Server perspective) are hosted as block volumes—iSCSI and/or Fibre-Channel-connected LUNs and file systems as SMB and/or NFS volumes. High availability starts with the network connectivity supporting the storage and server interconnectivity. Any design for the storage infrastructure should avoid single points of failure. Because many white papers and publications cover storage-area networking and network-attached storage resilience, those topics are not covered in detail in this paper.
    [Show full text]
  • IBM Cloud Object Storage System On-Premises Features and Benefits Object Storage to Help Solve Terabytes-And-Beyond Storage Challenges
    IBM Cloud Cloud Object Storage Data Sheet IBM Cloud Object Storage System on-premises features and benefits Object storage to help solve terabytes-and-beyond storage challenges The IBM® Cloud Object Storage System™ is a breakthrough platform Highlights that helps solve unstructured data challenges for companies worldwide. It is designed to provide scalability, availability, security, • On-line scalability that offers a single manageability, flexibility, and lower total cost of ownership (TCO). storage system and namespace • Security features include a wide The Cloud Object Storage System is deployed in multiple range of capabilities designed to meet configurations as shown in Figure 1. Each node consists of Cloud security requirements Object Storage software running on an industry-standard server. • Reliability and availability characteristics Cloud Object Storage software is compatible with a wide range of of the system are configurable servers from many sources, including a physical or virtual appliance. • Single copy protected data with In addition, IBM conducts certification of specific servers that geo-dispersal creating both efficiency customers want to use in their environment to help insure quick and manageability initial installation, long-term reliability and predictable performance. • Compliance enabled vaults for when compliance requirements or locking down data is required Data source Multiple concurrent paths ACCESSER® LAYER ......... Multiple concurrent paths SLICESTOR® LAYER ......... Figure 1: Multiple configurations of
    [Show full text]
  • CIFS/NFS) Administrator's Guide
    Hitachi Data Ingestor File System Protocols (CIFS/NFS) Administrator's Guide Product Version Getting Help Contents MK-90HDI035-13 © 2013- 2015 Hitachi, Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or stored in a database or retrieval system for any purpose without the express written permission of Hitachi, Ltd. Hitachi, Ltd., reserves the right to make changes to this document at any time without notice and assume no responsibility for its use. This document contains the most current information available at the time of publication. When new or revised information becomes available, this entire document will be updated and distributed to all registered users. Some of the features described in this document might not be currently available. Refer to the most recent product announcement for information about feature and product availability, or contact Hitachi Data Systems Corporation at https://portal.hds.com. Notice: Hitachi, Ltd., products and services can be ordered only under the terms and conditions of the applicable Hitachi Data Systems Corporation agreements. he use of Hitachi, Ltd., products is governed by the terms of your agreements with Hitachi Data Systems Corporation. Hitachi is a registered trademark of Hitachi, Ltd., in the United States and other countries. Hitachi Data Systems is a registered trademark and service mark of Hitachi, Ltd., in the United States and other countries. Archivas, Essential NAS Platform, HiCommand, Hi-Track, ShadowImage, Tagmaserve, Tagmasoft, Tagmasolve, Tagmastore, TrueCopy, Universal Star Network, and Universal Storage Platform are registered trademarks of Hitachi Data Systems Corporation.
    [Show full text]
  • Comptia A+ Acronym List Core 1 (220-1001) and Core 2 (220-1002)
    CompTIA A+ Acronym List Core 1 (220-1001) and Core 2 (220-1002) AC: Alternating Current ACL: Access Control List ACPI: Advanced Configuration Power Interface ADF: Automatic Document Feeder ADSL: Asymmetrical Digital Subscriber Line AES: Advanced Encryption Standard AHCI: Advanced Host Controller Interface AP: Access Point APIPA: Automatic Private Internet Protocol Addressing APM: Advanced Power Management ARP: Address Resolution Protocol ASR: Automated System Recovery ATA: Advanced Technology Attachment ATAPI: Advanced Technology Attachment Packet Interface ATM: Asynchronous Transfer Mode ATX: Advanced Technology Extended AUP: Acceptable Use Policy A/V: Audio Video BD-R: Blu-ray Disc Recordable BIOS: Basic Input/Output System BD-RE: Blu-ray Disc Rewritable BNC: Bayonet-Neill-Concelman BSOD: Blue Screen of Death 1 BYOD: Bring Your Own Device CAD: Computer-Aided Design CAPTCHA: Completely Automated Public Turing test to tell Computers and Humans Apart CD: Compact Disc CD-ROM: Compact Disc-Read-Only Memory CD-RW: Compact Disc-Rewritable CDFS: Compact Disc File System CERT: Computer Emergency Response Team CFS: Central File System, Common File System, or Command File System CGA: Computer Graphics and Applications CIDR: Classless Inter-Domain Routing CIFS: Common Internet File System CMOS: Complementary Metal-Oxide Semiconductor CNR: Communications and Networking Riser COMx: Communication port (x = port number) CPU: Central Processing Unit CRT: Cathode-Ray Tube DaaS: Data as a Service DAC: Discretionary Access Control DB-25: Serial Communications
    [Show full text]
  • Evaluation of Active Storage Strategies for the Lustre Parallel File System
    Evaluation of Active Storage Strategies for the Lustre Parallel File System Juan Piernas Jarek Nieplocha Evan J. Felix Pacific Northwest National Pacific Northwest National Pacific Northwest National Laboratory Laboratory Laboratory P.O. Box 999 P.O. Box 999 P.O. Box 999 Richland, WA 99352 Richland, WA 99352 Richland, WA 99352 [email protected] [email protected] [email protected] ABSTRACT umes of data remains a challenging problem. Despite the Active Storage provides an opportunity for reducing the improvements of storage capacities, the cost of bandwidth amount of data movement between storage and compute for moving data between the processing nodes and the stor- nodes of a parallel filesystem such as Lustre, and PVFS. age devices has not improved at the same rate as the disk ca- It allows certain types of data processing operations to be pacity. One approach to reduce the bandwidth requirements performed directly on the storage nodes of modern paral- between storage and compute devices is, when possible, to lel filesystems, near the data they manage. This is possible move computation closer to the storage devices. Similarly by exploiting the underutilized processor and memory re- to the processing-in-memory (PIM) approach for random ac- sources of storage nodes that are implemented using general cess memory [16], the active disk concept was proposed for purpose servers and operating systems. In this paper, we hard disk storage systems [1, 15, 24]. The active disk idea present a novel user-space implementation of Active Storage exploits the processing power of the embedded hard drive for Lustre, and compare it to the traditional kernel-based controller to process the data on the disk without the need implementation.
    [Show full text]
  • Samba-3 by Example
    Samba-3 by Example Practical Exercises in Successful Samba Deployment John H. Terpstra May 27, 2009 ABOUT THE COVER ARTWORK The cover artwork of this book continues the freedom theme of the first edition of \Samba-3 by Example". The history of civilization demonstrates the fragile nature of freedom. It can be lost in a moment, and once lost, the cost of recovering liberty can be incredible. The last edition cover featured Alfred the Great who liberated England from the constant assault of Vikings and Norsemen. Events in England that finally liberated the common people came about in small steps, but the result should not be under-estimated. Today, as always, freedom and liberty are seldom appreciated until they are lost. If we can not quantify what is the value of freedom, we shall be little motivated to protect it. Samba-3 by Example Cover Artwork: The British houses of parliament are a symbol of the Westminster system of government. This form of government permits the people to govern themselves at the lowest level, yet it provides for courts of appeal that are designed to protect freedom and to hold back all forces of tyranny. The clock is a pertinent symbol of the importance of time and place. The information technology industry is being challenged by the imposition of new laws, hostile litigation, and the imposition of significant constraint of practice that threatens to remove the freedom to develop and deploy open source software solutions. Samba is a software solution that epitomizes freedom of choice in network interoperability for Microsoft Windows clients.
    [Show full text]
  • Virtfs—A Virtualization Aware File System Pass-Through
    VirtFS—A virtualization aware File System pass-through Venkateswararao Jujjuri Eric Van Hensbergen Anthony Liguori IBM Linux Technology Center IBM Research Austin IBM Linux Technology Center [email protected] [email protected] [email protected] Badari Pulavarty IBM Linux Technology Center [email protected] Abstract operations into block device operations and then again into host file system operations. This paper describes the design and implementation of In addition to performance improvements over a tradi- a paravirtualized file system interface for Linux in the tional virtual block device, exposing guest file system KVM environment. Today’s solution of sharing host activity to the hypervisor provides greater insight to the files on the guest through generic network file systems hypervisor about the workload the guest is running. This like NFS and CIFS suffer from major performance and allows the hypervisor to make more intelligent decisions feature deficiencies as these protocols are not designed with respect to I/O caching and creates new opportuni- or optimized for virtualization. To address the needs of ties for hypervisor-based services like de-duplification. the virtualization paradigm, in this paper we are intro- ducing a new paravirtualized file system called VirtFS. In Section 2 of this paper, we explore more details about This new file system is currently under development and the motivating factors for paravirtualizing the file sys- is being built using QEMU, KVM, VirtIO technologies tem layer. In Section 3, we introduce the VirtFS design and 9P2000.L protocol. including an overview of the 9P protocol, which VirtFS is based on, along with a set of extensions introduced for greater Linux guest compatibility.
    [Show full text]